Design Thinking for Oncology Bedflow

Design Thinking for Oncology Bedflow


Penn Medicine, Office of the Chief Medical Officer


The liquid oncology service of the hospital, which provided chemotherapy and other blood-cancer treatments, was severely delayed. Length of stay (LOS) and time of day of discharge was increasing. Thus, patients waiting for those beds were constantly being rescheduled, which was not only infuriating, but which could potentially effect dose intensity and dose regularity. The process was equally frustrating for physicians, who could not guarantee their patients a treatment date. Moreover, the health system couldn't serve as many patients as it could at an efficient capacity, which was resulting in a loss of revenue. The Chief Medical Officer asked our Innovation Center: How might we expedite this system to improve quality of care? He thought the solution might be to make residents more efficient in the discharge process. 

My Role

Innovation Design Lead:

  • User research

  • Data analysis

  • Problem definition

  • Root cause analysis

  • Inter-disciplinary discussions

  • Ideation of solutions

  • Prototyping

  • Executive Presentations


After weeks of user research with physicians, nurse practitioners, staff, and patients, I first created a journey map of the many possibilities for redesign. Opportunities ranged from the patient being scheduled through discharge back home. We presented this map in our first meeting with executive leadership.

The patient journey with opportunities for redesigns

In that first meeting, the C-Suite requested our team focus on the upfront scheduling process, which was a root of the many pain points.

Next, through more inquiry, we created a root-cause analysis. We discovered that the many causes of rescheduling were not due to the discharge process, but in fact that we were putting the wrong patients in these beds. Patients with lower acuity (less sick) were taking the beds of those who were more sick and truly needing them. Additionally, by working with the data center and analyzing treatment times, it became clear that we weren't maximizing the capacity over weekends nor preparing patients to be ready to receive treatment as soon as they received a bed.

The enormous root cause analysis of the problem space

A part of the root cause analysis

I mapped crowdsourced user ideas beneath the root cause analysis, to see how many qualities of solutions we could design into one or two robust prototypes

By crowdsourcing ideas, and looking at the metrics those ideas would improve, I designed two prototypes as our deliverable. First was a survey system that could gather data on mid-acuity patients that needed an out-patient sort of "halfway" service. This mid-level acuity clinic would serve patients in an out-patient setting, without unnecessarily squandering all the hospital beds. 

My second prototype was for a new calendar algorithm and protocal system, which would maximize weekend and evening capacity.

This survey was designed to test needs of a mid-acuity clinic idea before we actually spent resources building one

The steps of our prototype process for a mid-acuity clinic


These solutions were piloted with clinical staff, which revealed data about staffing a clinic, and patient needs.

These two prototypes were then developed into lasting implementations at Penn Medicine.

This design sprint was one of the first major projects of the Innovation Center, and showed how a user-centered design process could produce ideas as prototypes that could be tested quickly and inexpensively, before rolling out larger initiatives. We renamed the sprint from "Time to Discharge" to "Time to Chemo" to express how we actually reimagined the problem definition in order to solve the true root need.

We then created this "Design Sprint Service Menu" to see how much time similar projects would take 1 designer per year of time allocation.

Our design-sprint consulting service mapped out for one designer's allocation


Additional Links:

Read the Final Report shared with the executive team

Designing an Employee Hypertension Program

Designing an Employee Hypertension Program